An analysis method based on complementarity of a nucleic acid nucleotide sequence can analyze genetic traits directly. Accordingly, this analysis is a very powerful means for identification of genetic diseases, canceration, microorganisms etc. Further, a gene itself is the object of detection, and thus time-consuming and cumbersome procedures such as in culture can be omitted in some cases.
Nevertheless, the detection of a target gene present in a very small amount in a sample is not easy in general so that amplification of a target gene itself or its detection signal is necessary. As a method of amplifying a target gene, the PCR (polymerase chain reaction) method is known. (Science, 230, 1350-1354, 1985). Currently, the PCR method is the most popular method as a technique of amplifying nucleic acid in vitro. This method was established firmly as an excellent detection method by virtue of high sensitivity based on the effect of exponential amplification. Further, since the amplification product can be recovered as DNA, this method is applied widely as an important tool supporting genetic engineering techniques such as gene cloning and structural determination. In the PCR method, however, there are the following noted problems: a special temperature controller is necessary for practice; the exponential progress of the amplification reaction causes a problem in quantification; and samples and reaction solutions are easily contaminated from the outside to permit nucleic acid mixed in error to function as a template.
As genomic information is accumulated, analysis of single nucleotide polymorphism (SNPs) comes to attract attention. Detection of SNPs by means of PCR is feasible by designing a primer such that its nucleotide sequence contains SNPs. That is, whether a nucleotide sequence complementary to the primer is present or not can be inferred by determining whether a reaction product is present or not. However, once a complementary chain is synthesized in error in PCR by any chance, this product functions as a template in subsequent reaction, thus causing an erroneous result. In practice, it is said that strict control of PCR is difficult with the difference of only one base given at the terminal of the primer. Accordingly, it is necessary to improve specificity in order to apply PCR to detection of SNPs.
On one hand, a method of synthesizing nucleic acid by a ligase is also practically used. The LCR method (ligase chain reaction, Laffler T G; Garrino J J; Marshall R L; Ann. Biol. Clin. (Paris), 51:9, 821-6, 1993) is based on the reaction in which two adjacent probes are hybridized with a target sequence and ligated to each other by a ligase. The two probes could not be ligated in the absence of the target nucleotide sequence, and thus the presence of the ligated product is indicative of the target nucleotide sequence. Because the LCR method also requires control of temperature for separation of a complementary chain from a template, there arises the same problem as in the PCR method. For LCR, there is also a report on a method of improving specificity by adding the step of providing a gap between adjacent probes and filling the gap by a DNA polymerase. However, what can be expected in this modified method is specificity only, and there still remains a problem in that control of temperature is required. Furthermore, use of the additional enzyme leads to an increase in cost.
A method called the SDA method (strand displacement amplification) [Proc. Natl. Acad. Sci. USA, 89, 392-396, 1992] [Nucleic Acid. Res., 20, 1691-1696, 1992] is also known as a method of amplifying DNA having a sequence complementary to a target sequence as a template. In the SDA method, a special DNA polymerase is used to synthesize a complementary chain starting from a primer complementary to the 3′-side of a certain nucleotide sequence while displacing a double-stranded chain if any at the 5′-side of the sequence. In the present specification, the simple expression “5′-side” or “3′-side” refers to that of a chain serving as a template. Because a double-stranded chain at the 5′-side is displaced by a newly synthesized complementary chain, this technique is called the SDA method. The temperature-changing step essential in the PCR method can be eliminated in the SDA method by previously inserting a restriction enzyme recognition sequence into an annealed sequence as a primer. That is, a nick generated by a restriction enzyme gives a 3′-OH group acting as the origin of synthesis of complementary chain, and the previously synthesized complementary chain is released as a single-stranded chain by strand displacement synthesis and then utilized again as a template for subsequent synthesis of complementary chain. In this manner, the complicated control of temperature essential in the PCR method is not required in the SDA method.
In the SDA method, however, the restriction enzyme generating a nick should be used in addition to the strand displacement-type DNA polymerase. This requirement for the additional enzyme is a major cause for higher cost. Further, because the restriction enzyme is to be used not for cleavage of both double-stranded chains but for introduction of a nick (that is, cleavage of only one of the chains), a dNTP derivative such as α-thio dNTP should be used as a substrate for synthesis to render the other chain resistant to digestion with the enzyme. Accordingly, the amplification product by SDA has a different structure from that of natural nucleic acid, and there is a limit to cleavage with restriction enzymes or application of the amplification product to gene cloning. In this respect too, there is a major cause for higher cost. In addition, when the SDA method is applied to an unknown sequence, there is the possibility that the same nucleotide sequence as the restriction enzyme recognition sequence used for introducing a nick may be present in a region to be synthesized. In this case, it is possible that a complete complementary chain is prevented from being synthesized.
NASBA (nucleic acid sequence-based amplification, also called the TMA/transcription mediated amplification method) is known as a method of amplifying nucleic acid wherein the complicated control of temperature is not necessary. NASBA is a reaction system wherein DNA is synthesized by DNA polymerase in the presence of target RNA as a template with a probe having a T7 promoter added thereto, and the product is formed with a second probe into a double-stranded chain, followed by transcription by T7 RNA polymerase with the formed double-stranded chain as a template to amplify a large amount of RNA (Nature, 350, 91-92, 1991). NASBA requires some heat denaturation steps until double-stranded DNA is completed, but the subsequent transcriptional reaction by T7 RNA polymerase proceeds under isothermal conditions. However, a combination of plural enzymes such as reverse transcriptase, RNase H, DNA polymerase and T7 RNA polymerase is essential, and this is unfavorable for cost similarly to SDA. Further, because it is complicated to set up conditions for a plurality of enzyme reaction, this method is hardly widespread as a general analytical method. In the known reactions of amplification of nucleic acid, there remain problems such as complicated control of temperature and the necessity for plural enzymes as described above.
For these known reactions of synthesizing nucleic acid, there are few reports on an attempt for further improving the efficiency of synthesis of nucleic acid without sacrificing specificity or cost. For example, in a method called RCA (rolling-circle amplification), it was shown that single-stranded DNA having a series of nucleotide sequences complementary to a padlock probe can be synthesized continuously in the presence of a target nucleotide sequence (Paul M. Lizardi et al., Nature Genetics, 19, 225-232, July, 1998). In RCA, a padlock probe having a special structure wherein each of the 5′- and 3′-terminals of a single oligonucleotide constitutes an adjacent probe in LCR is utilized. Then, the continuous reaction of synthesizing complementary chain with the padlock probe as a template which was ligated and cyclized in the presence of a target nucleotide sequence is triggered by combination with a polymerase catalyzing the strand displacement-type reaction of synthesizing complementary chain. Single-stranded nucleic acid having a structure of a series of regions each consisting of the same nucleotide sequence is thus formed. A primer is further annealed to this single-stranded nucleic acid to synthesize its complementary chain and a high degree of amplification is thus realized. However, there still remains the problem of the necessity for a plurality of enzymes. Further, triggering of synthesis of complementary chain depends on the reaction of ligating two adjacent regions, and its specificity is basically the same as in LCR.
For the object of supplying 3′-OH, there is a known method in which a nucleotide sequence is provided at the 3′-terminal with a sequence complementary thereto and a hair pin loop is formed at the terminal (Gene, 71, 29-40, 1988). Synthesis of complementary chain with a target sequence itself as a template starts at the hairpin loop to form single-stranded nucleic acid composed of the complementary nucleotide sequence. For example, a structure in which annealing occurs in the same chain at the terminal to which a complementary nucleotide sequence has been linked is realized in PCT/FR95/00891. In this method, however, the step in which the terminal cancels base pairing with the complementary chain and base pairing is constituted again in the same chain is essential. It is estimated that this step proceeds depending on a subtle equilibrium state at the terminal of mutually complementary nucleotide sequences involving base pairing. That is, an equilibrium state maintained between base pairing with a complementary chain and base pairing in the same chain is utilized and the only chain annealing to the nucleotide sequence in the same chain serves as the origin of synthesis of a complementary chain. Accordingly, it is considered that strict reaction conditions should be set to achieve high reaction efficiency. Further, in this prior art, the primer itself forms a loop structure. Accordingly, once a primer dimer is formed, amplification reaction is automatically initiated regardless of whether a target nucleotide sequence is present or not, and an unspecific synthetic product is thus formed. This can be a serious problem. Further, formation of the primer dimer and subsequent consumption of the primer by unspecific synthetic reaction lead to a reduction in the amplification efficiency of the desired reaction.
Besides, there is a report that a region not serving as a template for DNA polymerase was utilized to realize a 3′-terminal structure annealing to the same chain (EP713922). This report also has the same problem as in PCT/FR95/00891 supra in respect of the utilization of dynamic equilibrium at the terminal or the possibility of unspecific synthetic reaction due to formation of a dimer primer. Further, a special region not serving as a template for DNA polymerase should be prepared as a primer.
Further, in various signal amplification reactions to which the principle of NASBA described above is applied, an oligonucleotide having a hairpin structure at the terminal thereof is often utilized to supply a double-stranded promoter region (JP-A 5-211873). However, these techniques are not those permitting successive supply of 3′-OH for synthesis of a complementary chain. Further, a hairpin loop structure having a 3′-terminal annealed in the same chain is utilized for the purpose of obtaining a DNA template transcribed by RNA polymerase is utilized in JP-A10-510161 (WO96/17079). In this method, the template is amplified by using transcription into RNA and reverse transcription from RNA to DNA. In this method, however, the reaction system cannot be constituted without a combination of a plurality of enzymes.